Major histocompatibility complex class II association and induction of T cell responses by carbohydrates and glycopeptides

Conclusions In the following study we have investigated whether the inability to generate T cell immunity against carbohydrate antigens is the result of a failure of carbohydrates to form a complex with class II MHC molecules or a deficiency of carbohydratespecific T cells in the immune repertoire. These possibilities were examined by, first, determining the MHC binding capacity of pure carbohydrate molecules and, second, by assessing whether carbohydrate-specific T cells could be induced against a glycosylated T cell peptide epitope which had the capacity to interact with class II MHC restriction elements.In the first set of experiments, a group of synthetic and natural oligosaccharides and glycolipids were tested for their ability to bind IA^d class II molecules. Of the 26 carbohydrates analyzed, none were found to bind significantly to IA^d. Although only a small number of carbohydrate molecules of limited structural heterogeneity was tested for MHC binding, it appears that pure carbohydrates and glycolipids do not have the appropriate structures necessary to bind class II MHC molecules.Since carbohydrates may inherently lack MHC binding activity, we next addressed whether glycopeptides having MHC binding activity could induce T cells specific for the carbohydrate structure. A series of analogs of a well-characterized T cell determinant, the chicken OVA 323–339 peptide, was synthesized with a N-acetyl glucosamine substitution placed within the peptide molecule. Experiments performed with the glycosylated OVA peptides indicated three general types of response. In the first type of response, the addition of GlcNAc-Asn to the OVA peptide completely destroyed MHC binding capacity and, as a consequence, these peptides were nonimmunogenic for T cells in vivo. Peptides substituted at A329, A330, and A332 all showed this pattern. The Ala in position 332 was previously shown to be involved in interacting with the MHC molecule; thus, residues with bulky side chain groups would be expected to interfere with binding [22]. For positions A329 and A330, the lack of MHC binding could not be explained by the bulkiness of the GlcNAc side chain. For these peptides, the sugar molecule may have caused an undefined perturbation in peptide structure which completely ablated MHC binding.In the second category of response, substitution of GlcNAc-Asn appeared to be well tolerated since such analogs bound class II MHC molecules and primed T cells in vivo. This pattern was observed when the N-acetyl glucosamine was placed at an MHC contact site (e.g., at V327) or at a location sufficiently outside the core region where it was less likely to affect MHC binding and interact with the T cell receptor (e.g., at positions Q325 and N335). The cells generated against this set of peptides were highly cross-reactive for the nonglycosylated analog, further emphasizing that T cell recognition was not directed at the carbohydrate residue.A third type of response was observed when GlcNAc-Asn was placed near or within the core region of the peptide in positions A326, H328, and H331. The MHC binding capacity of these peptides was substantially but not completely diminished and the residual amount of binding present after glycosylation was still sufficient for the glycopeptide to be immunogenic in H-2^d mice. Interestingly, T cells raised against these glycopeptides failed to cross-react with its respective Asn-substituted analog, suggesting that the carbohydrate structure on the peptide was an integral part of the antigenic determinant recognized by the T lymphocyte.Collectively, results obtained in the present study indicate that the lack of T cell immunity against carbohydrate molecules is not due to a hole in the T cell repertoire, but is more likely due to the inability of carbohydrates to associate with MHC molecules with sufficient affinity. The failure of carbohydrate molecules to prime T cells may be circumvented by substituting sugar residues at strategic locations within an immunogenic T cell epitope in such a way that MHC binding is not drastically affected and T cell receptor recognition is possible. Indeed, under these conditions, our studies demonstrate that T cells with specificity for the carbohydrate moiety can be generated. It may be of interest to determine the effect of such glycopeptides on the humoral immune response of an otherwise T-independent antigen and to further examine the MHC interaction of natural and synthetic glycopeptides containing sugar residues of higher complexity in conjunction with the fine specificity of T cells induced with such carbohydrate molecules.